Table 4 Assessment of the model performance on the test set. Values in parentheses represent 95% confidence intervals
From: GPS driving: a digital biomarker for preclinical Alzheimer disease
Model inputs | Precision | Recall | F1 score | AUC |
|---|---|---|---|---|
Age and APOE ε4 status | 0.84 (0.802–0.875) | 0.79 (0.770–0.861) | 0.85 (0.833–0.852) | 0.88 (0.861–0.927) |
Driving features | 0.89 (0.862–0.917) | 0.76 (0.716–0.796) | 0.82 (0.794–0.840) | 0.82 (0.782–0.932) |
Driving features and age | 0.94 (0.909–0.959) | 0.84 (0.794–0.876) | 0.88 (0.858–0.906) | 0.94 (0.881–0.962) |
Driving features, age, and APOE ε4 status | 0.96 (0.939–0.981) | 0.88 (0.837–0.912) | 0.91 (0.893–0.937) | 0.96 (0.903–0.981) |